Rare earth oxides: Pr-Nd oxide FOB was quoted at $115-127/kg, down $1/kg; Pr-Nd oxide CIF Rotterdam was quoted at $108-122/kg, down $1/kg. Cerium oxide FOB was quoted at $2,005-2,007/mt, holding steady for now; cerium oxide CIF Rotterdam was quoted at $2,722-2,792/mt, holding steady for now. Praseodymium oxide FOB was quoted at $126-130/kg, holding steady for now. Neodymium oxide FOB was quoted at $141-153/kg, down $10/kg. Terbium oxide FOB was quoted at $1,213-1,301/kg, up $3/kg.
Rare earth metals: Pr-Nd alloy FOB was quoted at $136-142/kg, down $1/kg; Pr-Nd alloy CIF Rotterdam was quoted at $140-144/kg, down $1/kg. Lanthanum metal FOB was quoted at $3/kg. Praseodymium metal FOB was quoted at $158-166/kg. Neodymium metal FOB was quoted at $148-160/kg, holding steady for now. Terbium metal FOB was quoted at $1,515-1,599/kg, up $3/kg.
Trading: Some FOB quotes were lowered under the influence of domestic prices, but with limited rigid demand, transactions were hard to conclude. Heavy rare earth prices remained high, with quotes maintained solely due to policy factors.
Recycling process: HPSR laboratory validation
A team at the Max Planck Institute for Sustainable Materials (MPI-SusMat) led by Rafael Gitti Tortoretto Fim, in collaboration with Matic Jovičević-Klug, Dierk Raabe, Fraunhofer IWKS, and Brazil's IPEN, released a preprint (not peer-reviewed) on ChemRxiv on September 2, 2026: At a 10-gram laboratory scale, hydrogen plasma smelting reduction (HPSR) processed scrap from end-of-life NdFeB hard disk magnets for 2.5 minutes, reducing oxygen content by up to 66%; uncoated scrap after melt spinning achieved a maximum energy product of 72.4 kJ/m³, compared with 46.5 kJ/m³ for material derived from untreated scrap, an improvement of approximately 56%, with no virgin rare earths added throughout the process; nickel coating dissolved into the melt and degraded magnetic properties, highlighting the necessity of scrap pretreatment. Simulated costs were €11,500-15,500/mt, with economics still pending pilot-scale validation.
Mining and beneficiation pilots: Vietnam's Dong Pao and Greenland's Sarfartoq
Hanoi approved an 18-month semi-industrial pilot under Decision No. 4657/QD-UBND, with Mining Technology and Construction Consulting JSC operating a 100 kg/hour production line using a patented Vietnamese flotation-hydrometallurgy process to process 100 mt of Dong Pao ore into mixed rare earth oxides with TREO >95%. The output is prohibited from sale and serves only for technical validation, with no separation of individual magnet rare earths. South Korea's KIGAM is simultaneously cooperating with Vietnam on beneficiation and leaching of Dong Pao ore. Greenland Minerals purchased the Sarfartoq carbonatite rare earth project from Neo Performance Materials for $35 million. The ST1 deposit accounts for 1% of the 191 km² exploration license, with five additional mineral occurrences identified along a 32 km mineralized belt. An independent preliminary assessment based on NI 43-101 resources and 23,000 meters of drilling showed a pre-tax NPV of $2.05 billion and an IRR of 118.6%. ST1 is planned to produce Pr-Nd oxide at approximately one-third of China's ex-China output annually. The company plans infill drilling, pilot-scale beneficiation and metallurgy, and a pre-feasibility study, with the field team already back on site.
Western magnet supply chain: defense demand and recycling capacity coming online
Sprott reported that Pentagon demand for specialty rare earth magnets could reach 10,000 mt/year by 2030 (a threefold increase from current levels), absorbing nearly half of non-Chinese production. Non-Chinese NdFeB magnet production is 20,000-25,000 mt/year, versus approximately 250,000 mt globally. NATO listed rare earths among 12 defense-critical raw materials and committed to raising defense spending to 5% of GDP by 2035. USSM and Ionic Rare Earths USA signed a 50:50 joint venture term sheet on September 3, replacing the November 2025 MOU, to build a recycling plant at a 1,750-acre site in Fredericktown, Missouri. USSM will contribute $95 million in CAPEX, while Ionic will license its 10 mt/year Belfast pilot technology, with each party contributing $5 million in FEED costs, processing NdFeB/SmCo and heavy rare earths, with a target to convert to a definitive agreement by the end of 2026. Cycle Materials raised $75 million in strategic financing (total equity of $237 million), with ERI participating and supplying feedstock. Its McBee plant in South Carolina, costing $82 million, will expand processing capacity from 2,000 mt/year of magnetic material to 6,000 mt/year, and MREO production from 600 mt to 1,800 mt, with commissioning in 2028. Its Mesa plant in Arizona began operations in Q3 to process 25,000 mt of retired components. It secured a 10-year exclusive scrap rights agreement with Vacuumschmelze's Sumter plant in South Carolina, an offtake agreement with Solvay for MREO, and an offtake agreement with Glencore for recovered copper. Neo Performance Materials signed an offtake/processing agreement with France's Carester: Neo will receive dysprosium/terbium oxides from Caremag (Lacq, commissioning in Q1 2026) for its Narva plant in Estonia (which shipped 18,000 magnet samples in April 2025, with Phase 1 capacity expanding from 2,000 mt/year to 5,000 mt/year, commercial in H2 2026), while Neo will in turn supply NdFeB scrap to Carester for recovery of Pr-Nd/dysprosium/terbium for recycling back, and will toll-process Carester's mixed rare earth carbonates, retaining light rare earths and returning heavy rare earths. Lindian Resources signed a 10+5+5-year offtake agreement with Carester, under which Carester will purchase 70% of heavy rare earth carbonates from Kazakhstan's SARECO for supply to Lacq.
Technology R&D and resource growth
JAEA developed a rare earth adsorbent usable in both water and oil, with functional groups enabling targeted binding. In experiments, europium was recovered from water-oil mixtures and desorbed for regeneration using strong acid. Nexon Geochem (India) signed an MOU with Russia's Giredmet to cooperate on rare earth deep processing and NdFeB magnet technology, planning to build an integrated oxide-to-magnet plant in Hyderabad, targeting 120,000 mt/year NdFeB capacity by FY2033. Case Western Reserve University received a grant from the US Department of Energy's Critical Minerals office (one of seven projects totaling $10 million), led by Rohan Akolkar, to use patented molten salt electrolysis technology to reduce costs and improve efficiency in producing heavy rare earth metals. Defense Metals' Wicheeda project in British Columbia: hole WI26-96 intersected 20.3 meters at 6.1% TREO from 61 meters, and hole WI26-93 intersected 184 meters at 1.7% TREO from 91.6 meters (including 57 meters at 2.5%). Brazilian Rare Earths Limited (BRE) withdrew its original Rocha da Rocha plan and adopted a standalone Monte Alto + Camaçari plan, with post-tax NPV of $6 billion, IRR of 90%, and a payback period of 1.1 years, excluding Sulista inferred resources.
Supply-demand periods and regional collaboration
Vietnam's Dong Pao pilot, Greenland's Sarfartoq, Sarfartoq ST1, Wicheeda, and Monte Alto constitute samples of non-Chinese resource-side growth; USSM-Ionic, Cycle Materials, Neo-Carester-Lindian, JAEA adsorbent, and MPI-SusMat HPSR constitute samples of recycling and separation closed loops; KIGAM-Vietnam, Nexon-Giredmet, and JOGMEC-Iwatani-Carester Lacq constitute samples of regional collaboration. End-user validation remains incomplete for all magnet recycling economics, HPSR ton-scale costs, Dong Pao single rare earth separation, and Sarfartoq pre-feasibility study.
![Pr-Nd Prices Raised in the Rare Earth Market, Trading in Stalemate with Strong Wait-and-See Sentiment [SMM Rare Earth Daily Review]](https://imgqn.smm.cn/usercenter/THqPM20251217171743.jpeg)
